Retroviral vectors are currently the most efficient system for gene transfer into hematopoietic stem cells (HSC) for gene therapy. In prior studies, we have shown expression at high levels in mice transplanted with bone marrow carrying standard retroviral vectors which use the transcriptional regulatory elements of Moloney murine leukemia virus (MoMuLV). However, if the marrow is serially transplanted into a second generation of recipient mice, non-expression from the vector ("silencing") is frequently seen. The lack of expression is associated with increased levels of cytosine methylation in the vector LTR. Vector silencing has also been seen by other investigators, using different cell types. Therefore, investigations to understand the mechanisms of vector silencing and to identify vectors which will overcome silencing are of great importance to the successful application of gene therapy to treat a variety of diseases, including genetic diseases, cancer and AIDS. Our hypothesis is that cis- acting sequences in or near the MoMuLV LTR inhibit gene expression, leading to DNA methylation which maintains the expression inactivity. Removal of negatively-acting cis-elements or prevention of methylation will permit persistent vector expression. To test this hypothesis, we will extend our previous analysis of the methylation of cytosine residues in the MoMuLV LTR which occurs in association with transcriptional inactivity. We will then use complementary techniques to inhibit DNA methylation in murine gene transfer/BMT recipients and determine the effects on vector expression. We will also use a series of novel retroviral vectors which we have developed to contain alterations in the cis-acting elements of the LTR which permit increased expression in embryonal stem cells. These new vectors will be examined for their expression activity in murine HSC, defined by studies with serial transplantation of transduced bone marrow and by direct analysis of expression of a surface antigen reporter gene in HSC defined by immunophenotypic properties. In all, these studies will provide new information concerning the mechanisms by which gene expression by retroviral vectors is silenced and lead to identification of vector designs which permit more persistent expression.